Responsive image

Malmenator

Network Based Intelligent Malware Detection

Malmenator Reinventing Malware Protection

Malmenator introduces a smart way for network protection from malwares.
Our light weight hardware backed by a powerful software can be connected to any server on a network to offer complete protection from malwares

  • We secure your network in 3 ways:
  • Network Level: We scan each network packet to track & eliminate malicious behaviours
  • PC Level: We can identify the risk level of each PC on the network for effective malware detection
  • Flow Level: Our Machine Learning + Rule Based detection system can identify malicious network traffic and flows

Responsive image

Why Malmenator?


Easy to Use

No Setup Needed
Just plug in our hardware to a server in the network and enjoy great security

Smart Scanning

Say NO to scanning everything
Unlike traditional antivirus softwares we only scan suspicious flows thanks to our scanning optimisation procedures

Intelligent Malware Detection

Machine Learning is our strength
We combine rule based malware identification with Machine Learning for robust malware detection

Cost Effective

We are a one time investment
We aim to eliminate standard subscriptions and costs that come along with antivirus solutions

Let's talk Tech

Raspberry Pi 4 Model B based hardware packs our software into a single hardware component which can be connected to any server on the network for malware detection
Responsive image
We make use of Snort, an open source packet sniffer, and write our own protocols to effectively analyse the network for malwares
Responsive image
We employ a tensorflow based custom estimator for classifying the probability of having a malware for any given network flow.
We combine this with a signature based malware detection method to ensure maximum accuracy
Responsive image
We use an elasticsearch along with Kibana to visualise live network packets and risks for each PC on the network
Our custom dashboard on top of Kibana also provides a method to look into the infected files and take necessary actions
Responsive image

Timeline

September 2019
Project Plan & Research

1. Background Research on Malwares, Network Analysis, Antivirus & Machine Learning in Malware Detection
2. Assess Project Scope and Milestones

October 2019
Research on Network Analysis
1. Developing on Snort for detecting Malwares on Network
2. Creating a sample Backdoor Malware for testing
November 2019
Implement Hardware

1. Working on packaging Network Analysis on Raspberry Pi
2. Test Malware Detection on Network using Raspberry Pi

December 2019
Implement Dashboard

1. Work on ELK Stack to visualize data from Network Analysis
2. Develop a one stop Dashboard for the entire product

January 2020
Interim Report & Presentation

1. Consolidate a mid-term report for the project
2. Present the project to Supervisor and second examiner

February 2020
Malware Detection & Risk Analysis

1. Develop a Machine Learning model to classify malicious network flows
2. Analyse risk of each PC for node based risk classification

March 2020
Optimize Network Flow Scanning

1. Optimize network flow scanning by removing redundancies
2. Narrow down scope of scan to potentially risky flows

April 2020
Production & Testing

1. Deploy the project on cloud and Raspberry Pi
2. Test the project on various viruses for performance

May 2020
Project Exhibition

1. Create a Project Poster for exhibition
2. Test the project for the exhibition

Meet Us

Piyush Jha

Year 4
Major: Computer Science
Minor: Finance
University of Hong Kong

David Han

Year 4
Major: Computer Science
Minor: Math
University of Hong Kong

Our Mentor

Dr. Dirk Schneiders

PhD
Computer Science
Lecturer
University of Hong Kong

Presentation & Reports

...

Presentation

Open
...

Project Plan

Open
...

Final Engineering Report

Open
...

Final Research Report

Open
...

Interim Report

Open
...

Poster

Open
...

Project Video

Open